CLINICAL ARTICLEJ Neurosurg 126:1951–1958, 2017

IntraoperatIve monitoring of motor evoked potentials (MEPs) by transcranial electrical stimulation (TES) is useful for motor function preservation.16–18,23 Howev-

er, the accuracy of MEP monitoring during supratentorial neurosurgery is insufficient.4,19 Lower-extremity transcra-nial MEPs (LE-tMEPs) have been widely applied, espe-cially in the field of spinal surgery; however, LE-tMEP monitoring during supratentorial surgery was found to have a success rate of only 66%.1 This low success rate could be attributed to several factors. The distance be-tween electrodes on the scalp and the motor cortex of the LE is a major factor. Cortical MEPs (cMEPs) are con-sidered more reliable than tMEPs, and Maruta et al.5 re-ported that cMEPs are useful for LE-tMEP monitoring.7 Although cMEP monitoring by the direct cortical stimu-

lation method offers advantages over the tMEP method, the placement of subdural electrodes on the motor cortex of the LE is difficult or impossible when the craniotomy location is far from the motor cortex.

We usually perform tMEP monitoring by using con-stant current electrical stimulation. Our experience shows that LE-tMEPs often require electrical current stimula-tion using large currents (> 120 mA) and can cause pa-tient movement because of excessive muscle contraction around the head. This movement causes shaking of the mi-croscopic view and disturbs the operation. Moreover, dur-ing TES, most of the current spreads laterally through the scalp because of the skull’s high resistance; only a small percentage (approximately 20%) of the current seems to pass into the brain.22 There is no doubt that the reduction

ABBREVIATIONS CMAP = compound muscle action potential; cMEP = cortical MEP; LE = lower extremity; MEP = motor evoked potential; TES = transcranial electrical stimulation; tMEP = transcranial MEP; UE-tMEP = upper-extremity tMEP.SUBMITTED March 13, 2016. ACCEPTED July 11, 2016.INCLUDE WHEN CITING Published online September 23, 2016; DOI: 10.3171/2016.7.JNS16643.

Effects of transcranial stimulating electrode montages over the head for lower-extremity transcranial motor evoked potential monitoringRyosuke Tomio, MD, Takenori Akiyama, MD, PhD, Takayuki Ohira, MD, PhD, and Kazunari Yoshida, MD, PhD

Department of Neurosurgery, School of Medicine, Keio University, Tokyo, Japan

OBJECTIVE The aim of this study was to determine the most effective electrode montage to elicit lower-extremity trans-cranial motor evoked potentials (LE-tMEPs) using a minimum stimulation current.METHODS A realistic 3D head model was created from T1-weighted images. Finite element methods were used to visualize the electric field in the brain, which was generated by transcranial electrical stimulation via 4 electrode montage models. The stimulation threshold level of LE-tMEPs in 52 patients was also studied in a practical clinical setting to de-termine the effects of each electrode montage.RESULTS The electric field in the brain radially diffused from the brain surface at a maximum just below the electrodes in the finite element models. The Cz-inion electrode montage generated a centrally distributed high electric field with a current direction longitudinal and parallel to most of the pyramidal tract fibers of the lower extremity. These features seemed to be effective in igniting LE-tMEPs.Threshold level recordings of LE-tMEPs revealed that the Cz-inion electrode montage had a lower threshold on average than the C3-C4 montage, 76.5 ± 20.6 mA and 86.2 ± 20.6 mA, respectively (31 patients, t = 4.045, p < 0.001, paired t-test). In 23 (74.2%) of 31 cases, the Cz-inion montage could elicit LE-tMEPs at a lower threshold than C3-C4.CONCLUSIONS The C3-C4 and C1-C2 electrode montages are the standard for tMEP monitoring in neurosurgery, but the Cz-inion montage showed lower thresholds for the generation of LE-tMEPs. The Cz-inion electrode montage should be a good alternative for LE-tMEP monitoring when the C3-C4 has trouble igniting LE-tMEPs.https://thejns.org/doi/abs/10.3171/2016.7.JNS16643KEY WORDS transcranial motor evoked potential; transcranial electric stimulation; lower extremity; neurosurgery; electrode montage; diagnostic and operative techniques

©AANS, 2017 J Neurosurg Volume 126 • June 2017 1951

Unauthenticated | Downloaded 07/03/20 04:51 AM UTC

R. Tomio et al.

J Neurosurg Volume 126 • June 20171952

of lateral current spread by minimizing the TES current is important for LE-tMEP monitoring during neurosurgery.

However, the electric field generated by TES has not been well studied and cannot be easily estimated from in vivo studies. Accordingly, to investigate efficient TES methods for LE-tMEP monitoring, we visualized the elec-tric field in the brain during tMEP monitoring by using realistic finite element models made from standard MR images. Several authors have described the usefulness of this method for evaluating electrical fields in the brain during therapeutic transcranial direct current stimulation.6 Holdefer et al. performed a 2D study of tMEPs,3 Stecker performed a 3D study of tMEPs using a simple 3-layered spherical model,14 and we have described the electric field generated by TES for intraoperative tMEP monitoring during frontotemporal craniotomy using a 3D realistic head model.20

The aim of this study was to identify the electrode montage most effective in eliciting LE-tMEPs using a minimum stimulation current. To accomplish this, we in-vestigated both a computer simulation by finite element models and the stimulation threshold level of LE-tMEPs in 52 patients in a prospective study.

MethodsModel Simulation StudyRealistic 3D Head Model Creation and Electrode Placement

The realistic 3D head models created in this study were developed from International Consortium for Brain Map-

ping (ICBM) T1-weighted images obtained from Brain-Web (http://brainweb.bic.mni.mcgill.ca/). Image process-ing and segmentation of 1 × 1 × 1–mm3 resolution im-ages were done using ScanIP and +ScanFE (version 7.0, Simpleware Ltd.). The brain, CSF, skull, subcutaneous fat, and skin layer were obtained from the T1-weighted im-ages. These finite element models meshed into more than 1.5 × 107 tetrahedral elements (that is, more than 1.5 × 106 df) for the 1 × 1 × 1–mm3 resolution head models.

Next we created 4 models of electrode montages: C3-C4 model, C1-C2 model, Cz-inion model, and Cz-forehead model (Fig. 1). Details of the electrode design and placement were the same as in our previous publica-tion.20 Corkscrew-type bipolar electrodes were placed on the scalp at sites based on the international 10–20 system in the C3-C4 model and C1-C2 model. The anode was placed on the left and the cathode on the right in these models. Electrodes were placed at sites Cz and inion in the Cz-inion model and at Cz and the midline point near the hairline in the Cz-forehead model. Although some au-thors insist that a Cz-Fz electrode montage is efficient for LE-tMEPs, we placed the electrodes in a more anterior position than Fz because the distance between Cz and Fz was too close. The Cz electrode was the anode in both the Cz-inion and Cz-forehead models.

Tissue Conduction Properties and Calculations and Analyses by Comsol Multiphysics

All tissue layers were modeled as homogeneous and isotropic with respect to electrical conductivity and per-

FIG. 1. All 4 models of electrode montages are shown: C3-C4, C1-C2, Cz-inion, and Cz-forehead. Locations of the anode (red circles) and cathode (green circles) are indicated as well. Figure is available in color online only.

Unauthenticated | Downloaded 07/03/20 04:51 AM UTC

Electrode montages for lower-extremity transcranial MEP monitoring

J Neurosurg Volume 126 • June 2017 1953

mittivity based on data from the Human organs property database for computing simulation (http://cfd-duo.riken.go.jp/cbms-mp/index.htm). Details of the conductivity values were the same as in our previous publication,20 and the values were not the same but were nearly consistent with prior reports.3,6 The finite element mesh exported by ScanIP was read into Comsol Multiphysics (version 5.1, Comsol AB). Details of the finite element mesh generation by ScanIP and electric field calculation by Comsol Mul-tiphysics were the same as in our previous publication.20

We focused on the electric field distribution plotted on the coronal and sagittal cross-sections and the brain surface in each model. We made coronal sections con-taining both the electrodes and the motor cortex and pro-duced sagittal sections containing the motor cortex of the LE nearest to the interhemispheric fissure (midline). The magnitude of the electric field to the brain surface was estimated in surface studies of each model. All surface studies of the electric field were smoothed by replacing the value at each node with the average values in the neigh-boring triangles. The electric field values of cross-section studies were visualized by a color scale with a range from 0 V/m (blue) to 30 V/m (red; Fig. 2).

The Z-axis vector compartment of the electric field was also studied in coronal and sagittal sections. The Z-axis was defined as perpendicular to the axial plane and paral-lel to the vertical direction during the standing position,

and the X-axis was defined as perpendicular to the sagittal plane. The Z-axis was considered to be almost perpendic-ular to the pyramidal tract fibers. The electric field values of the Z-axis vector compartment around the midline of the brain containing the pyramidal tract were visualized using a color scale with a range from 0 V/m (blue) to 30 V/m (red; Fig. 3).

The electric line of force (equal to the electric current line) was pictured as red lines in the coronal and sagittal sections of the electric field. These red lines only represent the direction of the electric current. The magnitude of the current is not represented by these lines (Fig. 4).

Limitations of the Finite Element ModelsThe limitation details regarding the use of an isotropic

layer, the direct current application mode by Comsol Mul-tiphysics, and a mesh generation are described in our previ-ous publication.20

We had planned to use a whole-brain model that in-cluded the entire brain as well as levels of the foramen magnum, using 1 × 1 × 1–mm3 resolution images through-out the study at first, but our computer memory (32 GB RAM) was insufficient to perform the finite element method calculation. We, therefore, developed the supra-tentorial models, which included the telencephalon, at a 1 × 1 × 1–mm3 resolution to downsize the elements of the

FIG. 2. The electric fields of all electrode montage models are shown. From the left, the 3D model view, coronal section, sagittal section, and surface view of the electric field are shown for each model. The color scale ranges from 0 V/m (blue) to 30 V/m (red). Figure is available in color online only.

Unauthenticated | Downloaded 07/03/20 04:51 AM UTC

R. Tomio et al.

J Neurosurg Volume 126 • June 20171954

models. Head truncation artificially increased the electric field in the brain, especially near the inion electrode; how-ever, the truncation had a small effect on the LE motor cortex and pyramidal tract.

Clinical Validation by Recordings of the LE-tMEP ThresholdPatient Population

Intraoperative LE-tMEP monitoring and recording of the stimulation threshold were applied prospectively in a total of 52 consecutive patients, 27 male and 25 female, ranging in age from 19 to 81 years (mean 56.3 years). All patients underwent craniotomy procedures for aneurysm (9 patients) or tumor (43 patients). Any patients with motor weakness before surgery and/or a motor cortex or pyra-midal tract compressed by tumor were excluded. Usually intraindividual comparison between the C3-C4 and Cz-inion electrode montages was performed during fronto-temporal craniotomy, and comparison between the C3-C4 and Cz-forehead electrode montages was studied during occipital craniotomy cases.

Neurophysiological Monitoring Methods and Recording SettingsThe corkscrew electrode placements at C1, C2, C3, C4,

and Cz were based on the international 10-20 system. A forehead electrode was placed just posterior to the hairline at the midline of the head anterior to Fz, and an inion elec-trode was placed on the skin just over the inion.

We performed 3 combinations of intraindividual com-parisons between 2 electrode montages by recording LE-tMEP thresholds after the induction of general anesthesia: 1) a C3-C4 versus C1-C2 comparison in 9 patients, 2) a C3-C4 versus Cz-inion comparison in 31 patients, and 3) a C3-C4 versus Cz-forehead comparison in 12 patients. Each of the 3 combinations was performed independently. As noted above, comparisons between C3-C4 and C1-C2 were performed in only 9 cases and those between C3-C4 and Cz-forehead were performed in 12 cases. We then stopped performing these comparisons, as both the C1-C2 and Cz-forehead montages seemed to need higher electric currents than those applied in the C3-C4 montage to elicit an LE-tMEP.

All recording was performed before skin incision. We

FIG. 3. The Z-axis component distributions of the electric field for each electrode montage model are shown. From the top, the 3D model view, coronal section, and sagittal section are shown for each model. The color scale ranges from 0 V/m (blue) to 30 V/m (red). Figure is available in color online only.

Unauthenticated | Downloaded 07/03/20 04:51 AM UTC

Electrode montages for lower-extremity transcranial MEP monitoring

J Neurosurg Volume 126 • June 2017 1955

used an MEE-1200 series intraoperative monitoring sys-tem (Nihon Kohden Co. Ltd.) as the electrophysiological device. Anodal electrical constant current stimulation was performed with short trains of 5 stimuli consisting of rect-angular pulses with a pulse duration of 0.5 msec and an interstimulus interval of 2 msec. Compound muscle ac-tion potentials (CMAPs) of the abductor hallucis muscle were recorded as LE-tMEPs using sticking electrodes. The CMAPs of the abductor pollicis brevis muscle were also recorded as upper-extremity (UE)–tMEPs to confirm the validity of the monitoring system. A bandpass filter was set from 30 to 3000 Hz. The stimulation threshold was judged as a level of stimulation current (mA) that con-stantly elicited CMAPs > 50 μV. We limited the maxi-mum stimulation to 120 mA for safety because high stim-ulation levels often cause excessive muscle contraction of the head and neck by the lateral current and can stress the patient under head-pin fixation. Although both UE-tMEPs and LE-tMEPs can usually be monitored at less than 100 mA of TES, a value of 120 mA was recorded when tMEPs could not be ignited by 120 mA of stimulation. Analysis of the data by t-test was performed with SPSS Statistics 21 (IBM Corp.). Clinical data recording of this study was in accordance with the ethical standards of the institutional review board of Keio University and the Helsinki Declara-tion of 1975, as revised in 2000 and 2008. Informed con-sent to undergo neurophysiological monitoring was also obtained before surgery.

AnesthesiaAnesthesia was induced with a bolus of propofol and

remifentanil and was maintained with propofol and remi-fentanil at an average dosage. A short-acting muscle relax-ant was given as a bolus for intubation purposes. Although we did not use the “train of 4 stimuli” technique to test the level of relaxation because of time constraints prior to surgery, we routinely used neostigmine as a reversal agent for muscle relaxation after intubation and before recording the LE-tMEP threshold.

ResultsElectric Field Distribution in Each Electrode Montage in the Finite Element Models

The electric fields in all 4 models of electrode mon-tages were studied in 1 × 1 × 1–mm3 resolution models and are shown using the coronal and sagittal sections and the brain surface (Fig. 2). The electric field in the brain ra-dially diffused from the brain surface at a maximum just below the electrodes in the coronal sections. High electric fields were observed on both hand motor areas just under the electrodes in the C3-C4 model. The LE motor cortex areas seemed to be more strongly stimulated in the C1-C2 model, but the high electric field distribution and its depth were limited to a narrower area in the C1-C2 model than in the C3-C4 model.

There was also a high electric field in the LE motor

FIG. 4. The electric field and current direction (electric line of force) for all of the electrode montage models are shown. From the left, the 3D model view, coronal section, and sagittal section are shown for each model. The color scale ranges from 0 V/m (blue) to 30 V/m (red), and the direction of the electric current is indicated by red lines. Note that these red lines represent only the elec-tric force and the direction of the electric field, not the magnitude of the current. Figure is available in color online only.

Unauthenticated | Downloaded 07/03/20 04:51 AM UTC

R. Tomio et al.

J Neurosurg Volume 126 • June 20171956

area in both the Cz-inion and Cz-forehead models, es-pecially the Cz-inion model, which showed broader and deeper high electric field areas around the midline of the brain than those in the other 3 models. The brain areas around the inion were very strongly stimulated in the Cz-inion model, but these high electric fields were affected by the “truncation effect” and would hardly affect the pyra-midal tract because these high electric field areas were far from the pyramidal tract.

The Z-axis vector component of the electric field is considered to be almost parallel to most pyramidal tract fibers from the LE motor cortex area in the white matter (Fig. 3). There was a very small high Z-axis vector com-ponent in the brain around the midline in the C3-C4 and C1-C2 models. Conversely, there was a high Z-axis vector component around the midline in the Cz-inion and Cz-forehead models.

The line of electric force connecting the anode and cathode in each model was studied (Fig. 4). The line was almost parallel to the X-axis in the C3-C4 model around the midline of the brain and was bent in an arc in the C1-C2 model. Both the Cz-forehead and the Cz-inion models had a vertical line of electric force almost parallel to the Z-axis under the anode. The Cz-inion model had more verti-cal currents around the motor cortex of the LE than the Cz-forehead, which had obliquely disposed current lines in an anteroposterior direction.

Threshold of LE-tMEP Stimulation in Each Electrode Montage in Patients

The Cz-inion electrode montage had a lower LE-tMEP threshold on average than the C3-C4 montage, a statisti-cally significant difference in the comparison of 31 pa-tients (76.5 ± 20.6 mA and 86.2 ± 20.6 mA, respectively, p < 0.001, paired t-test; Table 1). The LE-tMEP latency on average was not statistically significantly different be-tween the Cz-inion and the C3-C4 montages (29 patients, 46.1 ± 3.67 msec and 45.3 ± 2.96 msec, respectively, p < 0.195, paired t-test).

The Cz-inion could elicit LE-tMEPs at a lower thresh-old than the C3-C4 model in 23 (74.2%) of 31 cases. A stimulus of 120 mA ignited LE-tMEPs in all of the Cz-inion cases but could not ignite LE-tMEPs in 2 (6.4%) of 31 C3-C4 cases. However, the C3-C4 electrode montage could ignite UE-tMEPs at obviously lower stimulation lev-els on average than the Cz-inion (31 patients, 42.4 ± 13.1 mA and 93.5 ± 26.0 mA, respectively, p < 0.001, paired t-test). By 120 mA, the C3-C4 montage ignited UE-tMEPs in all cases, but 8 (25.8%) of the Cz-inion cases were not ignited. The C3-C4 montage could elicit UE-tMEPs at a lower threshold than the Cz-inion in all cases.

The Cz-forehead electrode montage had a higher LE-tMEP threshold than the C3-C4 arrangement, which reached statistical significance when a comparison of the 12 patients was performed (97.8 ± 27.1 mA and 85.2 ± 20.8 mA, respectively, p < 0.020, paired t-test). The Cz-forehead electrodes could elicit LE-tMEPs at the lower threshold in only 2 (16.7%) of 12 cases. The threshold of UE-tMEPs was also lower with the C3-C4 electrodes than with the Cz-forehead electrodes.

The C1-C2 electrode montage apparently had a sig-

nificantly higher threshold for LE-tMEPs than the C3-C4 electrode arrangement in all cases (9 patients, p < 0.001, paired t-test). Threshold values of UE-tMEPs were also higher in the C1-C2 than the C3-C4 montage (9 patients, p < 0.001, paired t-test). The average values of these elec-trode montages were 106.9 ± 17.9 mA and 83.6 ± 12.0 mA in LE-tMEPs and 66.7 ± 9.4 mA and 44.7 ± 16.1 mA in UE-tMEPs, respectively.

DiscussionRecording tMEPs has become a standard method of

monitoring motor function during neurosurgery, but the accuracy of LE-tMEP monitoring during supratentorial surgery is still insufficient. We have therefore discussed several theoretical and technical points of LE-tMEPs to speculate on more efficient TES methods for LE-tMEP ignition.

Anodal stimulation is a popular TES method for tMEPs and has been considered more effective than cathodal stimulation. This finding was first reported by Fritsch and Hitzig, followed by many others.2,10,12 Some reports have indicated the usefulness of cathodal stimulation for single neurons9,15 and the efficacy of biphasic stimulation;21 how-ever, anodal stimulation is useful enough and has remained the standard. Thus, an anodal electrode was usually placed over a target point of the motor cortex in the present study.

The electrical conductivity of TES and the intensity of the electric field in the brain would impact the ignition of tMEPs. Thus, electrode placement just over a target point of the brain seems to be logical given that the distance between the electrode and the targeted motor cortex and the direction of the electric current affects the efficacy of TES for tMEP ignition.

Changes in the electric field gradients along the direc-tion of the axon, which are called “activating functions” by Rattay,11 are considered more important than the magni-tude of the electric field in the static state.8,14 Peak depolar-ization is expected to be located where the activating func-tion attains its maximal value. We tried to simulate the activating function distribution, but our results may not ex-actly represent it (data not shown). However, the activating function is not the solution for a transmembrane potential change, but only its initial rate of change. Transmembrane

TABLE 1. Threshold values of LE- and UE-tMEPs in each electrode position*

Comparison LE-tMEP UE-tMEP

C3-C4 vs C1-C2 C3-C4 83.6 ± 12.0 44.7 ± 16.1 C1-C2 106.9 ± 17.9 66.7 ± 9.4C3-C4 vs Cz-inion C3-C4 86.2 ± 20.6 42.4 ± 13.1 Cz-inion 76.5 ± 20.6 93.5 ± 26.0C3-C4 vs Cz-forehead C3-C4 85.2 ± 20.8 37.4 ± 11.1 Cz-forehead 97.8 ± 27.1 108.4 ± 17.3

* Values expressed in mA.

Unauthenticated | Downloaded 07/03/20 04:51 AM UTC

Electrode montages for lower-extremity transcranial MEP monitoring

J Neurosurg Volume 126 • June 2017 1957

potential changes over time during and after the stimulus interval and the transmembrane potential depend in some way on the entire function of the electric field.8 Thus, we suggest that the intensity of the electrical field and its vec-tor direction should also be considered for the develop-ment of effective TES methods.

The anode placed on the Cz position is near the mo-tor cortex (nearest is 1 cm behind the Cz) of the LE. The Cz-inion and Cz-forehead models had this Cz anode, and the maximum intensity of the electric field radiated out from the motor cortex of the LE in our models. The C1-C2 and Cz-forehead models also showed maximum intensity around the motor cortex of the LE, but the distribution of a high electric field was narrower in these models than in the C3-C4 and Cz-inion schemes. The C1-C2 and Cz-forehead needed higher stimulation currents to elicit LE-tMEPs. This shows that parameters in addition to electric field in-tensity are important.

The direction of the electric lines of force that are gen-erated between anode and cathode is important because neurons are most strongly influenced by the component of the electric field parallel to their trajectory. Rudin and Eisenman reported that if the electrodes had a transverse orientation to the axons instead of a longitudinal orienta-tion, 4–5 times as much current would be needed to depo-larize the axon.13 Both intensity and orientation of stimula-tion to the brain are considered critical factors for tMEP ignition. All of our models showed that the direction of the electric line of force started perpendicular to both the anode and cathode. As the direction of current in the brain depends on the electrode’s position, the electrode’s orien-tation is quite important.

The motor cortex of the LE was located at both the con-vexity and the interhemispheric fissure. These anatomical features make ignition of LE-tMEPs complicated because the direction of the fibers that run to the pyramidal tract from the LE motor cortex differs between the convexity region and the interhemispheric region. The existence of sulci and gyri of the motor cortex makes this situation more complicated. Thus, the electric current between the electrodes has different effects on different fibers, depend-ing on each fiber’s direction. Our results indicate that the Cz-inion model has the most vertical current parallel to the Z-axis compared with the other models despite the above limitations. These results indicate that the Cz-inion electrode positions have an advantage in igniting the long perpendicular part of the fibers of the pyramidal tract. The Cz-inion model also had a diffuse high Z-axis component of the electric field around the midline of the brain, an electric field that seemed to have considerable impact on ignition of the pyramidal tract fibers. These features of the Cz-inion placement may be the reason for its efficacy for LE-tMEP ignition. The Cz-forehead electrode was posi-tioned almost anteroposteriorly rather than vertically. This difference seemed to affect the threshold of LE-tMEPs.

The threshold recordings indicated that the Cz-inion electrode montage had a lower threshold for LE-tMEP ig-nition. Our results also showed that LE-tMEPs could not be ignited by 120 mA of stimulation in 6.4% of the C3-C4 cases; however, the Cz-inion could elicit LE-tMEPs in all cases. Accordingly, we suggest that the Cz-inion is useful

for LE-tMEP ignition, especially as an effective alterna-tive in cases in which the C3-C4 or C1-C2 cannot ignite LE-tMEPs. Theoretical reasons for Cz-inion usefulness would be its centrally distributed high electric field and the vertical direction of current as mentioned above. The latency of the Cz-inion was almost the same as that of C3-C4, and it showed that the location of ignition of the py-ramidal tract fibers was not deeper but almost the same. Another advantage of the Cz-inion is wider adaptability because this electrode position can be applied in most neurosurgeries, except the midline suboccipital approach, in which a skin incision often reaches the inion. The Cz-inion cathode arrangement, which was near the neck and paraspinal muscles, was theoretically prone to contrac-tions of extracranial muscle groups, and some cases actu-ally showed stronger contractions than the C3-C4 at each threshold level. Although it seemed to be a disadvantage of the Cz-inion, most of our LE-tMEP monitoring cases using the Cz-inion stimulation did not show hazardous head movements during surgery. Locational modification of the Cz-inion cathode to increase the distance between the cathode and the muscles was useful when the Cz-inion caused considerable head movements.

Upper-extremity tMEP is the most widely accepted in-traoperative motor function monitoring method, and the C3-C4 (C1-C2 is also popular) electrode montage is stan-dard for UE-tMEP monitoring. Our results indicate that the threshold of UE-tMEP ignition was lower with the C3-C4 montage than with the other 3 montages. Although the threshold of the C3-C4 montage was higher than that of the Cz-inion arrangement for LE-tMEP ignition, it was still able to induce LE-tMEPs and was useful in most (93.5%) cases. Thus, the C3-C4 electrode position must remain the standard for tMEP monitoring in neurosurgery, and the Cz-inion electrode position would be a good alternative for LE-tMEPs when the C3-C4 has trouble igniting LE-tMEPs.

The Cz-forehead montage led to a higher threshold and seemed to be less effective than the C3-C4 and Cz-inion arrangements in eliciting LE-tMEPs. The Cz-forehead montage showed higher localized electric fields in the pyr-amidal tract of the LE in our model. However, the direc-tion of its anteroposterior electric current and the distribu-tion of its superficial electric field may not be as effective for tMEP ignition.

The C1-C2 montage is popular, but its threshold for tMEPs was higher than that for the C3-C4 montage. The distance between the electrodes is shorter, and both elec-trodes are located near the vertex of the head. As a result, the electric field and the right-to-left electric current are concentrated in superficial structures, which may nega-tively affect tMEP ignition.

The results of our threshold recordings indicated that the Cz-inion montage was useful for LE-tMEP ignition, that both the C1-C2 and Cz-forehead electrode montag-es were not as effective as C3-C4 and Cz-inion, and that Cz-inion was probably more useful than C3-C4. We also discussed the theoretical background using the results of finite element method simulations on several electrode montage models. However, the theoretical consideration of TES is still insufficient, especially in neurophysiologi-cal and bioelectricity fields, and further investigation is

Unauthenticated | Downloaded 07/03/20 04:51 AM UTC

R. Tomio et al.

J Neurosurg Volume 126 • June 20171958

needed to establish the most appropriate TES methods for tMEP monitoring.

ConclusionsThe C3-C4 electrode montage is the standard for tMEP

monitoring in neurosurgery, but the Cz-inion electrode montage had a lower threshold for the generation of LE-tMEPs than the C3-C4 montage. Thus, the Cz-inion elec-trode montage should be a good alternative for LE-tMEP monitoring when the C3-C4 electrode montage has trouble igniting LE-tMEPs.

AcknowledgmentsWe M. thank Hashiguchi (Keisoku Engineering, Tokyo) for

his contributions to the technical support of Comsol Multiphysics. This work was supported by the Keio University Grant-in-Aid for Encouragement of Young Medical Scientists.

References 1. Chen X, Sterio D, Ming X, Para DD, Butusova M, Tong T, et

al: Success rate of motor evoked potentials for intraoperative neurophysiologic monitoring: effects of age, lesion location, and preoperative neurologic deficits. J Clin Neurophysiol 24:281–285, 2007

2. Hern JEC, Landgren S, Phillips CG, Porter R: Selective exci-tation of corticofugal neurones by surface-anodal stimulation of the baboon’s motor cortex. J Physiol 161:73–90, 1962

3. Holdefer RN, Sadleir R, Russell MJ: Predicted current densi-ties in the brain during transcranial electrical stimulation. Clin Neurophysiol 117:1388–1397, 2006

4. Lee JJ, Kim YI, Hong JT, Sung JH, Lee SW, Yang SH: Intraoperative monitoring of motor-evoked potentials for supratentorial tumor surgery. J Korean Neurosurg Soc 56:98–102, 2014

5. Maruta Y, Fujii M, Imoto H, Nomura S, Oka F, Goto H, et al: Intra-operative monitoring of lower extremity motor-evoked potentials by direct cortical stimulation. Clin Neurophysiol 123:1248–1254, 2012

6. Miranda PC, Mekonnen A, Salvador R, Ruffini G: The elec-tric field in the cortex during transcranial current stimulation. Neuroimage 70:48–58, 2013

7. Motoyama Y, Kawaguchi M, Yamada S, Nakagawa I, Nishimura F, Hironaka Y, et al: Evaluation of combined use of transcranial and direct cortical motor evoked potential monitoring during unruptured aneurysm surgery. Neurol Med Chir (Tokyo) 51:15–22, 2011

8. Plonsey R, Barr RC: Bioelectricity: A Quantitative Ap-proach, ed 3. New York: Springer, 2007, pp 208–209

9. Porter R: Focal stimulation of hypoglossal neurons in the cat. J Physiol 169:630–640, 1963

10. Ranck JB Jr: Which elements are excited in electrical stimu-lation of mammalian central nervous system: a review. Brain Res 98:417–440, 1975

11. Rattay F: Ways to approximate current-distance relations for electrically stimulated fibers. J Theor Biol 125:339–349, 1987

12. Rothwell J, Burke D, Hicks R, Stephen J, Woodforth I, Crawford M: Transcranial electrical stimulation of the motor cortex in man: further evidence for the site of activation. J Physiol 481:243–250, 1994

13. Rudin DO, Eisenman G: The action potential of spinal axons in vitro. J Gen Physiol 37:505–538, 1954

14. Stecker MM: Transcranial electric stimulation of motor path-ways: a theoretical analysis. Comput Biol Med 35:133–155, 2005

15. Stoney SD Jr, Thompson WD, Asanuma H: Excitation of pyramidal tract cells by intracortical microstimulation: effec-tive extent of stimulating current. J Neurophysiol 31:659–669, 1968

16. Szelényi A, Bueno de Camargo A, Flamm E, Deletis V: Neu-rophysiological criteria for intraoperative prediction of pure motor hemiplegia during aneurysm surgery. Case report. J Neurosurg 99:575–578, 2003

17. Szelényi A, Hattingen E, Weidauer S, Seifert V, Ziemann U: Intraoperative motor evoked potential alteration in intra-cranial tumor surgery and its relation to signal alteration in postoperative magnetic resonance imaging. Neurosurgery 67:302–313, 2010

18. Szelényi A, Langer D, Kothbauer K, De Camargo AB, Flamm ES, Deletis V: Monitoring of muscle motor evoked potentials during cerebral aneurysm surgery: intraoperative changes and postoperative outcome. J Neurosurg 105:675–681, 2006

19. Tanaka S, Tashiro T, Gomi A, Takanashi J, Ujiie H: Sensitiv-ity and specificity in transcranial motor-evoked potential monitoring during neurosurgical operations. Surg Neurol Int 2:111, 2011

20. Tomio R, Akiyama T, Horikoshi T, Ohira T, Yoshida K: Vi-sualization of the electric field evoked by transcranial electric stimulation during a craniotomy using the finite element method. J Neurosci Methods 256:157–167, 2015

21. Ukegawa D, Kawabata S, Sakaki K, Ishii S, Tomizawa S, Inose H, et al: Efficacy of biphasic transcranial electric stim-ulation in intraoperative motor evoked potential monitoring for cervical compression myelopathy. Spine (Phila Pa 1976) 39:E159–E165, 2014

22. Watanabe K, Watanabe T, Takahashi A, Saito N, Hirato M, Sasaki T: Transcranial electrical stimulation through screw electrodes for intraoperative monitoring of motor evoked potentials. Technical note. J Neurosurg 100:155–160, 2004

23. Zhou HH, Kelly PJ: Transcranial electrical motor evoked po-tential monitoring for brain tumor resection. Neurosurgery 48:1075–1081, 2001

DisclosuresThe authors report no conflict of interest concerning the materi-als or methods used in this study or the findings specified in this paper.

Author ContributionsConception and design: Tomio. Acquisition of data: Tomio. Anal-ysis and interpretation of data: Tomio. Drafting the article: Tomio, Akiyama. Critically revising the article: Tomio, Akiyama, Ohira. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Tomio. Statistical analysis: Tomio. Administrative/technical/mate-rial support: Tomio. Study supervision: all authors.

CorrespondenceRyosuke Tomio, Department of Neurosurgery, Keio University School of Medicine, 35 Shinanomachi, Shinjuku, Tokyo 160-8582, Japan. email: tomy0807@hotmail.com.

Unauthenticated | Downloaded 07/03/20 04:51 AM UTC